PROJECT SUMMARY Electroconvulsive therapy (ECT) is a highly effective treatment for severe depression and other mental health conditions, yet how ECT changes the brain to ameliorate the symptoms of depression is not well understood. Seizure generalization appears to be critical for successful antidepressant response to ECT, and stimulation dose is adjusted to ensure an ?adequate? seizure. Previous studies have linked seizure generalization during ECT with thalamocortical regions, and our own MRI studies indicate that thalamocortical networks exhibit neuroplasticity after treatment in patients who respond to ECT. Yet, despite recent growth in the number of MRI studies measuring long-term neuroplasticity across ECT sessions, relatively less is known about the neurobiology of seizure activity elicited during each session. The long-term goal of the proposed studies is to understand how ECT-induced seizures change brain networks to treat depression. To address this goal, we will take a data-driven, machine learning approach to pre-existing ECT-MRI datasets of patients with depression, to identify the neurobiological features measured with MRI (functional connectivity, gray-matter morphometry, electric fields) that optimally explain inter-patient variability in seizure indicators (seizure duration and stimulation dose). Aim 1 will identify biomarkers of susceptibility to ECT-induced seizures by analyzing pre-treatment MRI with respect to initial stimulation dose and seizure duration. Aim 2 studies will determine how brain network plasticity over the course of ECT relates to changes in stimulation dose and duration of ECT-induced seizures. These studies will further our understanding of therapeutic seizures in ECT, potentially facilitating prediction of optimal dose and/or treatment outcome to improve ECT and minimize side effects.